Topic: the difference of 20db

[spcyrfc]

this may be a silly question to some, but its something ive never really wrapped my head around...

ive been looking at getting a new pre, the aerco mp2 has options of 0-50 or 20-70 gain settings.

is there a way for me to quantify this difference in terms that i can understand? 
i used to run some marantz machines and they had a -20db pad... i can imagine what my levels look like with the pad on and pad off, but how do i translate that image into the 0-50 / 20-70 difference?

i have recorded in the past (with modded 660) in situations where with the pad off the signal was too hot, but with the pad on, it was to soft.  dont want to run into that situation after buying a new pre.
thanks
lb

[rustoleum]

for every increase of 10db, the human ear will perceive the sound to be twice as loud.  Thus, adding 20Db to your signal (assuming you can add that much before reaching full digital scale) would be perceived as 4 times louder than not.  Not sure if I answered your question, but that's certainly one difference between adding 20db and not.

[spcyrfc]

 thank you for the quantification.

so a pre with gain settings of 0-50 could handle a signal four times as loud as a pre with 20-70?

is the right way of looking at it?

and what is full digital scale, when is it a concern?

[Patrick]

thank you for the quantification.

so a pre with gain settings of 0-50 could handle a signal four times as loud as a pre with 20-70?

Microphone preamps are rated in the amount of amplification that it gives to a mic level (very low) signal.  70db amplification is hefty- you most likely will not be using that much for any type of audience recording.  So, to answer your question, a preamp that can amplify up to 70db can theoretically AMPLIFY a signal 4x as loud compared to preamp that is rated at 50db of gain max. 

I believe some of the Aerco's out there have a fixed 20db input gain stage.  I know the AD-20's have something like 17db as well. 

and what is full digital scale, when is it a concern?

Digital Full Scale refers to when you have used all of the available values given by a certain bit depth.  If you attempt to futher increase the level of the signal at DFS, you will get digital clipping.  This is (obviously) a concern when using digital systems, and avoiding artifacts as a result of digi clipping.

[SmokinJoe]

The real question is "how much gain do you need?"

I have a V3 and know that usually the gain on the V3 is set to something between 25-35 db for most shows that I tape.  It goes to 70db, and I can envision using 50db if I were micing an unamped solo acoustic situation (my friend's bluegrass band), but I can't envision using more that that, unless someday possibly to capture the loons at camp.

I tape mostly average rock bands, no super loud death-metal stuff.  If you do, I can see a situation where the minimum of 20db might be too much.

So... Most of the time for concert recording either will do.  But where do your occasional tendencies lie? 
- Do you see yourself at the superloud death metal show?  If so, go for the 0-50 model.
- Do you see yourself at the very quiet acoustic stuff, and maybe capturing nature sounds?  If so, go for the 20-70 model.

[Gutbucket]

The real question is "how much gain do you need?"

The loudness of what you are recording is the first factor in how much gain you'll need, but the second and just as important is the sensitivity of the mics you use.  Most of us are using powered condensers which are relatively sensitive (some more than others of course) and those don't require huge amounts of preamp gain to get good level to the ADC stage.  If you were to use dynamic mics, which are typically a good bit less sensitive than the average condenser, you'd need more preamp gain to achieve the same level.  One extreme example that applies to concert recording would be using ribbon mics to record something that isn't very loud like a string quartet, as ribbon mics are lower in sensitivity than most other dynamic mics and string quartets aren't very loud.  Yet that's a reasonable application that would probably require 50-70db of preamp gain.

If you are using condensers like most do here you'll probably be fine with the 0-50db range.

[DSatz]

spcyrfc, a preamp acts as an intermediary between your microphones and some kind of recording device. The inputs of any recording device will require a certain voltage in order to record at the levels you'd like to reach for the highest-level (generally, the loudest) sounds in the performance that you're recording.

You probably don't want to have to turn the input controls on your recorder ALL the way up, since that usually adds some amount of unnecessary noise. So the voltage that you want to send in to the recorder's inputs at the moments of maximum sound pressure level will depend on the design of the particular recorder, on where you choose to set the input level controls on that recorder, and on your personal choice of a maximum recording level relative to 0 dB (full scale).

Let's just pretend that you have recorder "X," and that its line inputs (when its record level controls have been set to some comfortable middle level) require exactly 1 Volt to reach a recording level of 1.5 dB below digital full scale. The recorder doesn't care how you produce that 1 Volt of signal--you can drive it from the line outputs of a stereo receiver, or the analog line outputs of another recorder, OR the outputs of a mike preamp among other possibilities. But given that recorder, that record level setting, and your desire to reach 1.5 dB below digital full scale, your goal is to feed a 1-Volt signal to the recorder during the absolute peak of the music you want to record. That will get you the record levels that you want, without adding undue noise.

OK. Now, the preamp is your tool for boosting the signals from your microphones exactly enough--not less, and not more--so that the level of those signals will be 1 Volt during the loudest sound(s) in the concert. Typically, a microphone might put out (say) 150 mV during loud sounds, and your preamp's mission is to boost that to 1 Volt. Thus the amount of gain you need from the preamp is whatever factor is required to boost what your microphones are actually putting out up to (in this case) 1 Volt.

At any given gain setting, the preamp will multiply the varying voltage at its input by some constant factor. If your microphones are sensitive enough to put out, say, 50 mV at the peak of the sound pressure during the concert, and your requirement is to get 1 Volt of signal out of the preamp at that same moment, then the preamp has to multiply the input voltage by a constant factor of 20, no? OK, a factor of 20 in voltage = about 26 dB of gain. If the mikes only put out 25 mV, then that multiplication of the voltage has to be by a constant factor of 40, or about 32 dB. If you set the preamp's gain up to 40 dB then it will multiply the (varying) input voltage by a constant factor of 100--it would raise a 10 mV signal to 1 Volt.

So it all depends on how much voltage is actually coming out of your microphones during the moments of the highest SPL, and how much voltage you would like to send in to the recorder at those moments. You set the preamp so that it supplies the needed gain for that situation. Do you see how this works? Your microphones have a constant sensitivity--for a certain sound pressure level, they put out a certain voltage. Your preamp then raises that voltage until it's suitable for your recorder's line inputs.

--best regards

[aaronji]

Stupid question, but:  you get those dB values from 20*log (ratio of voltages), right?  Thanks...

[DSatz]

aaronji, right. For the "voltage factor of 20" example that I used, log 20 = 1.3; for "voltage factor of 40," log 40 = 1.6; for "voltage factor of 100," log 100 = 2.0. Multiply each logarithm by 20 and you get 26, 32 and 40 dB respectively. (These are base-10 logarithms, rounded off of course.)

I mostly use an integrated preamp + A/D converter, which simplifies the picture considerably. For most of the live music recording that I do, with professional condenser microphones (sensitivity in the range of 10 mV/Pa), I find myself almost always setting my preamp gains in the range of 30 - 40 dB. It's remarkable how consistent that is.

With other types of microphones, of course, the range would be different, and if I had to feed different A/Ds or recorders with varying input sensitivities, that would add an additional variable to the equation. But due to my laziness and my fondness for ultra-quick, simple setups, I'm spared all that.

--best regards

[BC]

So... Most of the time for concert recording either will do.  But where do your occasional tendencies lie? 
- Do you see yourself at the superloud death metal show?  If so, go for the 0-50 model.
- Do you see yourself at the very quiet acoustic stuff, and maybe capturing nature sounds?  If so, go for the 20-70 model.

agree, with my mics I would run ~+35 at rock shows with a PA but for classical chamber music +55 which gave conservative levels. Not sure how the sensitivity of your mics compares though, the 4022's are relatively insensitive, I think 7mV/Pa off the top of my head.

[spcyrfc]

the Senns have a sensitivity of 20 mV/Pa (-34 dBV/Pa)
 
if the dpas have 7, wow, what a difference.

[DSatz]

spcyrfc, the difference between 20 mV/Pa and 7 mV/Pa is about 9.1 dB. If you're using these Sennheiser microphones, you'd select a preamp gain setting that's about 9.1 dB lower than what you would use with the DPA microphones for the same placement and sound pressure levels.

There's no free lunch, though, since each preamp has a limit on how high a voltage its input circuit can take without overload. Not every preamp will indicate (via LED or whatever) that its input circuit is being overdriven. Most such indicators are attached either to the output stage of the preamp, or to the stage that drives the output. A few years ago I tested one well-known portable preamp that (depending on various settings) could put out over 25% THD before its "Over" light came on. The distortion was due to input stage overload, as its output stage was faithfully amplifying all the distortion coming from the input stage.

The more sensitive your microphones are, the more vulnerable you are to that type of preamp overload. If you're using microphones with 20 mV/Pa sensitivity and you're recording loud music, you should check your preamp's overload limits very carefully--you may need to use resistive pads at the inputs of your preamp.

--best regards

[Myco]

DSatz always has good info to add. Good to have him around here.

Thanks guy! 

[spcyrfc]

DSatz always has good info to add. Good to have him around here.

Thanks guy! 

too true. still trying to find time to dissect/digest it all.

[carpa]

Thank you Dsatz for your clear explanation!
I thought that the more sensitive a mic is, the better for the preamp in order to use not too much gain (expecially for not-so-good preamps).
Today I received an Edirol R09HR; I played around a little bit and I think it is very good.
For better recordings I know I have to add something external (which I'll buy in the future), and I was leaning towards a phantom supply and something like:
a couple of AT's like 4021 (19,9 mV) or
Beyerdynamics (30 mV) or
New AT stereo BP 4025 (25 mV)
What I had in mind was using any of these sensitive mics in order to save gain of the internal Edirol preamp, or even trying to run them directly line-in for close miking (of course feeding them with phantom). I've read good things about the rode nt4 also, but I thought:less sensitive=not so good at the pourpose.

Now, what you said about overloading preamp input makes me wonder if I would have any problems with such a rig (I think that a 10 db pad would not do much...).
Can you ( or everybody else, of course)tell me your opinion, please? It would prove disappointing to choose mics, buy them
in order to have low noise and eventually have a "very low-noise distorted stuff!!!
c

[Gutbucket]

If you are not using the recorder to power the mics, you have a good bit of flexibility by being able to switch from using the mic input when you need more gain to using the line input when you need less gain.  But besides that gross adjustment of approximately 12db (maybe? guessing on the HR version) of sensitivity difference between in the two inputs, you also have additional control over input level built into the input level control on the machine.  That input control acts as either an attenuator or amplifier on the Edirol.  At high input level numbers (the number shown on the display when you adjust input level) you are amplifying the signal, but at low input numbers you are attenuating the input and reducing the signal level.  It acts just like adding an external attenuator (a variable one) before the input.

Of course at some point you could have an input level strong enough that you would still overload the input even using the line input with the level set as low as it goes (level 1, level 0 effectively mutes the input).  Fortunately the R-09 has a pretty high input level threshold and can take even strong line-level board patches and hot preamp outputs without distorting.  For specs concerning actual measured level differences between the inputs, the maximum input levels and preamp noise levels at various gain settings, check Guysonic's review and tests.

The point on that input level number scale where the R09 switches from attenuating the signal to adding gain is called the 'unity-gain' setting and is a popular subject of debate on this board.  Since I have older R-09s and the internal preamps and input setting numbering scale has changed on the HR I won't guess where that point is, and honestly it doesn't really matter as long as you get a feel for what is a good comfort range of input levels on the recorder given the mics you choose and the loudness of the sounds you record. 

[DSatz]

carpa, you wrote:

> I thought that the more sensitive a mic is, the better for the preamp in order to use not too much gain (expecially for not-so-good preamps).

With an extremely noisy preamp or an extremely quiet recording venue that could be true, but it's unlikely for most people here, I think.

Every microphone has a certain equivalent noise level, often 10 to 20 dB SPL or thereabouts. That amount of noise is always part of the signal which that microphone puts out. But microphone sensitivity and noise are tied together in a way that might not be obvious. Think this through, please: If you have two microphones with the same equivalent noise level and one microphone is twice as sensitive as the other, the more sensitive microphone will put out a residual noise voltage twice as high as the other microphone will put out.

It's crucial to be aware of that relationship. With a very noisy preamp, yes--you'd like to "swamp" its input noise with a nice, clean, strong signal from the microphone. But once your microphone is so sensitive that its own noise voltage equals or exceeds the preamp's input noise voltage, then you can't "swamp" the preamp's input noise any further--you'd be trying to cover noise with noise, and that doesn't work.

The same thing applies to room noise, and that's the real kicker. I'm a classical engineer primarily; classical recording sessions are very much affected by any noise from the room or even from outside the building you're recording in. Living in a big city, it's damned hard to find any recording venue quiet enough to let you hear the noise of a modern, professional condenser microphone. Most people here, from what I can see, record in places where the noise of their microphones AND the noise of any halfway reasonable preamp are both swamped by noise from the venue approximately 100% of the time.

If people here were into recording clavichords or certain nature sounds from a distance in ultra-quiet surroundings, using dynamic microphones (especially ribbons), it would be different. But people who make live recordings of amplified music with an audience present, using professional condenser microphones--preamp input noise just isn't going to be an audible factor unless your preamp is total crap. Not only do you not need to drown out that noise with extra oomph from your microphones, you can't do so, because your microphone's own signals are so full of room noise.

--best regards

P.S.: gutbucket, the difference in sensitivity between mike and line inputs on consumer equipment is typically 30 - 40 dB. Consumer "line level" is about 12 dB below pro "line level," but consumer "mike level" is even further below professional "mike level"--the mike inputs on consumer recorders are generally designed for use with low-cost dynamic microphones or electret condensers, and microphones in that class typically have 10, 15 or even 20 dB lower sensitivity than typical professional condenser microphones have. If the R-09 really had only a 12 dB difference between its mike and line inputs it would be quite unusual, and not at all practical for many users.

[taperwheeler]

Jumping in on the discussion, I am one who has been somewhat confused with understanding diff's between line and mic level, at least from an application perspective.  Mostly my concern is in regards to running a new setup consisting of low sensitivity modded at933 mics > ca9100 pre > MT2.  Ideally I would love to run the pre via the 1/4" line level inputs in most situations, however, when running via the 1/4" you cannot add any gain with the recorder, leaving just the 20db of gain available with the pre due to shortfalls of the recorder. I assume prior to any live testing that in some settings, the 20db won't give me enough oomph to reach an acceptable level on the recorder, example: in a quiet, acoustic performance.  Would it then make sense to run via the 3.5mm connection in those situations? 
From a previous thread/discussion with Guysonic he reported that the 3.5mm input measured "MT2 minijack mic maximum ~ -2.4dBu (the ZOOM H2 maximum LINE input!) "
Is there any way for me to figure out how much of a signal from my mics>pre will be too much for the 3.5mm input?

This technical stuff just confuses me.   

[carpa]

Thank you all.
DSatz, you've been very clear, and I appreciate your explanations very much. Being mic sensitivity and self-noise much related, the only real advantage in using a hi-sens mic would be in the possibility to go straight line-in without preamp, assuming that more or less the higher noise of the mic will be balanced with the less noise of the line input (which, on HR, has the possibility of changing the input level as well as mic-in).
Anyway, everybody is a "pro" in his field and a newbie in 100.000 others. I'm a professional musician and you are professionally involved in recording classical music, which is my work, so you can probably understand my needs better than others taping other music genres.
Apart from every technical consideration I would like any advice on the best way to complete my rig.
My R09HR is perfect for rehearsal and can be good even for taping a concert fot archive use or for internet posting (I can place it near the piano on a stand using it like a mic); I would know how I could  upgrade this without spending too much money.
I understand  that to make a big "jump" I'd have to spend a lot, and maybe it would prove nonsense not having the know-how to use gear, but I think that there must be a star-point. If somebody asked me an advice for buying a piano I wouldn say; "no way:a Steinway model D or, if you can't afford this, a digital keybord would suit all the rest being wasted money!"
I just suggested in my previous post a list of mikes (AT 4021, Beyers 930, the New AT stereo bp4025-nobody knows it yet-) chosen in terms of noise, sensitivity performance and price (max that I can afford). Even the idea of running mics to line in was intended in order to reduce noise and not else.
Given that, which mic choice would you do at my place? Compared to the stuff mentioned, in Italy I could get a Rode nt4 for 330Eeuros, but the samples I heard didn't convince me so much (a little too bright sound, and I suspect x-y not being the best for not too close micing- in case of ensembles), even I only heard guitars or pop ensembles.
thank you very much
I hope I can help someone with advices on pianos or classical music, so I don't feel so much abusing of your kindness
c

 

[Gutbucket]

Many here would argue that a cardioid pattern microphone is a better general application choice, but if you are primarily recording yourself playing piano in a decent sounding room with the freedom to place the mics where you'd like them, omnis could be a great choice. As you mention your application will be quite different than most people here.  Although you won't be 'close mic'ing' the instrument, you will have the freedom to place the mics much closer to the piano than the type of recording that most people here are doing from out in the audience or even at the edge of the stage.

Piano is a demanding instrument to record and sound natural, especially to the ear of the musician!  When recording piano with the option of placing the mics wherever you like, I find it hard to beat a pair of spaced omnis, regardless of price.  I'll also add that as a general rule I find omni mics sound more natural than directional mics, especially the models that are at a budget price point. It seems the more directional the mic pattern gets the more good sounding examples seem to cost. I also find that omnis are useful when the sound of the instrument 'in the room' is very important, such as in classical music.  Just my personal opinion.  [shrug]

[Gutbucket]

... P.S.: gutbucket, the difference in sensitivity between mike and line inputs on consumer equipment is typically 30 - 40 dB. Consumer "line level" is about 12 dB below pro "line level," but consumer "mike level" is even further below professional "mike level"--the mike inputs on consumer recorders are generally designed for use with low-cost dynamic microphones or electret condensers, and microphones in that class typically have 10, 15 or even 20 dB lower sensitivity than typical professional condenser microphones have. If the R-09 really had only a 12 dB difference between its mike and line inputs it would be quite unusual, and not at all practical for many users.

DS you are of course correct. I neglected to mention that the R-09 also has a sensitivity switch that adjusts between high and low mic sensitivity before the variable gain/attenuation adjustment.  The difference between the mic input set to 'high sensitivity' and the line input is indeed about 32db, but the difference of the mic input set to 'low sensitivity' and the line input is actually half my 12db guess at around 6db.  I'd switched input jacks once upon a time when my line-in jack broke and knew I didn't have to reduce my gain settings by more than ~10db so I just guessed 12 (thinking it surely couldn't be less  ). I so rarely use the high sensitivity mic input setting due to preamp noise that I neglected to mention that option.  In any case, although I try to keep the input setting in the R-09 away from either extreme by adjusting external preamp gain, the +16 dBv input limit figure below is a pretty healthy input level that probably won't require an input pad for this type of use.  I could be wrong on that though and I don't know what the distortion measurement would be.  I'm getting in a bit above my head here.

I dug out the info for the R-09 below, the HR version is probably similar-

R-09 input settings for 0dB VU FS-
 
                Setting@1(min)      Setting@30(max)
MIC HIGH        -16 dBv                 -45 dBv
MIC LOW        +10 dBv                 -20 dBv
LINE              +16 dBv                 -13 dBv 

(dBv = 0.775 volt RMS = 0 dB reference level)

Thanks for calling attention to that and making me think.

Respectfully,
GB

[chris319]

The SPL present at a microphone depends not only on how loud a performer is playing but also on the distance between the performer and the mic. Each doubling of the distance between performer and mic results in a 6 dB drop in amplitude (a 6 dB drop equals a halving of amplitude). If you can get a handle on the SPL at a given distance, you can get a handle on the amount of amplification needed to achieve a given output level:

http://www.miclisteningroom.org/gain.html

ribbon mics are lower in sensitivity than most other dynamic mics

Not necessarily. Ribbon mics have built-in step-up transformers which bring up their output levels. Compare these numbers:

Cascade Fathead: -56 dBV @ 1 Pa

AEA R84: -52 dBV @ 1 Pa

Royer R-121: -50 dBV @ 1 Pa

Shure SM58: -54.5 dBV @ 1 Pa

The ribbon mics are within a few dB of the SM58. Some are hotter.

[DSatz]

Chris, step-up transformers are pretty much a necessity in ribbon microphones. But they're far from being trivial components that you can take for granted--some are a lot better in quality than others, and in cheap ribbon mikes, you can bet that they're one place where major cost-cutting has been done. Good-sounding audio transformers--especially with significant voltage step-up--can't be made cheaply (at least, not so far).

Also, as transformers step up the output voltage, they also raise the output impedance. (Impedance is the ratio of voltage to current, and a step-up transformer essentially transforms current into voltage--whatever factor it multiplies the voltage is the factor by which it simultaneously divides the current.) Now, a ribbon microphone has much higher output impedance around its resonant frequency to begin with, and if a manufacturer chooses a transformer with a large voltage step-up to compensate for inherent lack of sensitivity, the output impedance around the resonant frequency can become so high that frequency-selective losses will occur at the input of the preamp, mixer or recorder. The microphone may therefore sound quite different depending on what equipment you use it with.

As an example, a certain high-quality U.S.-made ribbon microphone has a specified output impedance of 300 Ohms, measured at 1 kHz. That's already borderline at best; studio microphones shouldn't have an output impedance above 200 Ohms, because many preamp inputs are only about 1000 Ohms, and when that ratio is less than 5:1 you start getting noticeable losses from loading. But at this microphone's resonant frequency (which is near the low end of its range--below 100 Hz), its output impedance is well above 1000 Ohms--which means that more than half the signal at those frequencies will simply be thrown away at the inputs of many preamps.

This type of microphone has audibly better frequency balance (a major part of its very nice sound quality) when connected to a preamp with 5 to 10 kOhm input impedance. Such preamps, while not unheard of, certainly aren't the norm. If the step-up transformer in this microphone had used a more modest turns ratio, these loading effects would be correspondingly reduced, and the microphone would achieve its intended sound quality with a much wider range of preamps, mixers and recorders--though it would then have lower output voltage, and be less sensitive.

My point is, it pays to be a little wary of ribbon microphones that are almost as sensitive as conventional (moving-coil) dynamic microphones--you at least want to know how they got that way. I don't think I made the statement that you quoted, but I'll put things a different way: The dynamic microphones with the lowest sensitivity tend to be ribbon microphones with no active circuitry. But the professionals who prefer these microphones are willing to deal with the extra problems that they sometimes bring.

Among professional microphones you could take the Beyer M 130 or M 160 as examples, at 1 mV/Pa. This isn't any criticism of those microphones--they're way underappreciated as far as I'm concerned--but from personal experience, when I've tried to use them, I have more often run into radio frequency interference than I have with condenser microphones. In large part this is because the audio signal voltage in the cable was only about 1/10 what it would have been with a professional condenser microphone. That's a lot like amplifying the interference by a factor of 10.

--best regards

[Myco]

To add an actual experience of a simpleton to this discussion, I was at Ha Ha the Moose in Boston the other night getting ready to use my new Portico 5012 running into my HD-P2 via the mic inputs. I'm all set up and switched on in record pause mode. Even with the gain set at (+6db) on the Portico, and 0 on the P2 before the show was started my levels were running hot, about (-20db) just catching the PA and the crowd noise in the room. I was concerned that once the show started I'd be close to 0db if not clipping in this small room. My friend Joe mentioned that he recalled that there was a way to switch on a (-20db) pad on the P2 (it's right on the top of the deck actually). So I switched on the two channels for the (-20db) pad and it was golden, nice resulting recording and I averaged about (-6db) for the duration of the show. I ended up running the Portico at (+24db) and the P2 at about 3 on it's gain dial. Just a little non-technical input from a real-world application of the principles discussed here.

[carpa]

At this proposal, I wonder if, with a less pro gear like a pocket recorder and external mikes, the result could be achievedi by a 10 db pad on mikes plus the preamp input est to low-sensitivity on the recorder. I would say yes, but haven't tried yet.
c

[chris319]

many preamp inputs are only about 1000 Ohms

I had this discussion elsewhere not long ago. The consensus was that mic preamps are typically designed with input impedances of 10x the "typical" mic impedance, which is assumed to be 150 - 200 ohms; thus a "typical" preamp's input impedance would be around 1500 - 2000 ohms.